Help with Skew-t plots that require units in variables

[arunaskuciauskas]

HI:
I am working on developing a Skew-T plot where I want to include shading for CINH and CAPE. Using your MetPy Mondays #19 session, I am trying to calculate parcel_path:
parcel_path = mpcalc.parcel_profile(Pressure, TempC[0], Td[0])
At the bottom of this discussion is the error message that I receive, asking for units that I are not there. How do I provide the proper parameters for the parcel_path calculation? Thanks...

The attachment shows my progress. I am also working off of a Sounding file that is also attached.

WTEC_20190325_042301_edt.txt

import pandas as pd
import matplotlib.pyplot as plt
import metpy.calc as mpcalc
from metpy.plots import SkewT
from metpy.units import units
import numpy as np
fn = 'WTEC_20190325_042301_edt.txt'

# Read in RAOB, skip 1st 8 lines, grab the header data
df_title = pd.read_csv(fn, sep ='\s+', skiprows=5, nrows = 1)
df_title = str(df_title)

# Read in the rest of the file
df = pd.read_csv(fn, sep ='\s+', skiprows=8, nrows = 3800)

df_title_date = df_title[49:60]
print(df_title_date)
day = df_title_date[0:2]
month = df_title_date[3:5]
year = df_title_date[6:11]

# I couldn't perform an df.column name assignment, e.g., Lat = df['Lat'], so I applied indexing for each column
Hgt = df.iloc[:,1]
Pressure = df.iloc[:,2]
Temp = df.iloc[:,3]
RH = df.iloc[:,4]
Lat = df.iloc[:,7]
Lon = df.iloc[:,8]

wnd_dir = df.iloc[:,5]units('degree')
wnd_spd = df.iloc[:,6]units('knots')
rad = 4.0np.arctan(1)/180.
u_new = -wnd_spdnp.sin(radwnd_dir)
v_new = -wnd_spdnp.cos(rad*wnd_dir)

u = wnd_spdnp.cos(np.radians(wnd_dir))
v = wnd_spdnp.sin(np.radians(wnd_dir))

Convert Temperature to Celsius
TempC = Temp - 273.15

Calculate Td from RH
Td = Temp - ((100 - RH)/5.)

Change default to be better for skew-T
fig = plt.figure(figsize=(9, 11))

Initiate the skew-T plot type from MetPy class loaded earlier
skew = SkewT(fig, rotation=45)

Plot the data using normal plotting functions, in this case using
log scaling in Y, as dictated by the typical meteorological plot
skew.plot(Pressure, Temp, 'r')
skew.plot(Pressure, Td, 'g')

Set some appropriate axes limits for x and y
skew.ax.set_xlim(-30, 40)
skew.ax.set_ylim(1020, 100)

mask = Pressure >= 100

Add the relevant special lines to plot throughout the figure
skew.plot_dry_adiabats(t0=np.arange(233, 533, 10) * units.K,
alpha=0.25, color='orangered')
skew.plot_moist_adiabats(t0=np.arange(233, 400, 5) * units.K,
alpha=0.25, color='tab:green')
skew.plot_barbs(Pressure[mask][::75], u_new[mask][::75], v_new[mask][::75])

skew.plot_mixing_lines(Pressure=np.arange(1000, 99, -20) * units.hPa,
                   # linestyle='dotted', color='tab:blue')
Add some descriptive titles
SHIP = 'Ron Brown: ', Lat[1], Lon[1]
month=int(month)
day = int(day)
year = int(year)

dtg = month,'/',day,'/',year

table = str.maketrans(dict.fromkeys("()"))

dtg.translate(table)
plt.title(format(SHIP), loc='left')
plt.title(format(dtg), loc='right')

plt.xlabel('Temperature (C)')
plt.ylabel('Pressure hPa')

plt.show()

parcel_path = mpcalc.parcel_profile(Pressure, TempC[0], Td[0])

ValueError Traceback (most recent call last)
~\AppData\Local\Temp\10/ipykernel_1180/3592370993.py in
----> 1 parcel_path = mpcalc.parcel_profile(Pressure, TempC[0], Td[0])

~\Anaconda3\lib\site-packages\metpy\xarray.py in wrapper(*args, **kwargs)
1214
1215 # Evaluate inner calculation
-> 1216 result = func(*bound_args.args, **bound_args.kwargs)
1217
1218 # Wrap output based on match and match_unit

~\Anaconda3\lib\site-packages\metpy\units.py in wrapper(*args, **kwargs)
243 msg = (f'This function changed in {changed_version}--double check '
244 'that the function is being called properly.\n') + msg
--> 245 raise ValueError(msg)
246 return func(*args, **kwargs)
247

ValueError: This function changed in 1.0--double check that the function is being called properly.
parcel_profile given arguments with incorrect units: pressure requires "[pressure]" but given "none", temperature requires "[temperature]" but given "none", dewpoint requires "[temperature]" but given "none"
Any variable x can be assigned a unit as follows:
from metpy.units import units
x = units.Quantity(x, "m/s")

[DanielAdriaansen]

Per the error message:

Any variable x can be assigned a unit as follows:
from metpy.units import units
x = units.Quantity(x, "m/s")

I think you need to add units to Pressure, TempC, and Td before using them in the call to parcel_profile(). You can do this in various places in the code, but something like:

from metpy.units import units
TempC = units.Quantity(TempC,"degC")
Td = units.Quantity(Td,"degC")
Pressure = units.Quantity(Pressure,"hectopascals")
parcel_path = mpcalc.parcel_profile(Pressure, TempC[0], Td[0])

Please make sure the units I added are actually the units you're using.

[arunaskuciauskas]

Dan: Thanks, but the error message is the same after adding the units as you prescribed. I inserted the following into each variable: TempC = units.Quantity(TempC, "degC") Td = units.Quantity(Td,"degC") Pressure = units.Quantity(Pressure,"hectopascals") with the following error: TypeError Traceback (most recent call last)~\AppData\Local\Temp\11/ipykernel_14660/4184182665.py in <module> 23 24 # COnvert parameters into units---> 25 TempC = units.Quantity(TempC, "degC") 26 Td = units.Quantity(Td,"degC") 27 Pressure = units.Quantity(Pressure,"hectopascals") ~\Anaconda3\lib\site-packages\pint\quantity.py in __new__(cls, value, units) 255 def __new__(cls, value, units=None): 256 if is_upcast_type(type(value)):--> 257 raise TypeError(f"Quantity cannot wrap upcast type {type(value)}") 258 259 if units is None and isinstance(value, str) and value == "": TypeError: Quantity cannot wrap upcast type <class 'pandas.core.series.Series'> Should I assign the units into the variables before I read them in? Arunas Arunas Kuciauskas ***@***.*** 831-261-4342

…

On Thu, Dec 30, 2021 at 7:09 AM Dan Adriaansen ***@***.***> wrote: Per the error message: Any variable x can be assigned a unit as follows: from metpy.units import units x = units.Quantity(x, "m/s") I think you need to add units to Pressure, TempC, and Td before using them in the call to parcel_profile(). You can do this in various places in the code, but something like: from metpy.units import units TempC = units.Quantity(TempC,"degC") Td = units.Quantity(Td,"degC") Pressure = units.Quantity(Pressure,"hectopascals") parcel_path = mpcalc.parcel_profile(Pressure, TempC[0], Td[0]) Please make sure the units I added are actually the units you're using. — Reply to this email directly, view it on GitHub <#2279 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AQL3QFPSKNRPV24GLRXF43TUTRY23ANCNFSM5K7BY3JQ> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you authored the thread.Message ID: ***@***.***>

[arunaskuciauskas]

Dan:
Also, I read these variables in as strings. Do I need to convert them into some other type before trying the units function?

[DanielAdriaansen]

My apologies- I failed to notice that you had your data stored in Pandas objects. You'll need to convert the array of data from a Pandas Series object into another type that MetPy recognizes for unit operations prior to attaching the units.

Perhaps try this:

from metpy.units import units
TempC = units.Quantity(TempC.to_numpy(),"degC")
Td = units.Quantity(Td.to_numpy(),"degC")
Pressure = units.Quantity(Pressure.to_numpy(),"hectopascals")
parcel_path = mpcalc.parcel_profile(Pressure, TempC[0], Td[0])

Notice the addition of the to_numpy() method on the Pandas Series object. This converts from the Pandas Series object to a numpy object which the MetPy unit operations can work with.

If this doesn't work, then yes, I'd also double check that your numbers aren't stored as strings. Pandas read_csv() is pretty good at inferring data types, and so I think you should be OK.

[arunaskuciauskas]

Dan: Sorry that it took me so long to respond, I was ill. Using your approach above, please see below for the code and subsequent errors. I can't find a way in working on variables with units. Thanks, Arunas BTW: I'm new at this...If there's a more efficient way of presenting my issue, please let me know. ```python # I couldn't perform an df.column name assignment, e.g., Lat = df['Lat'], so I applied indexing for each column Hgt = df.iloc[:,1] # Pressure = df.iloc[:,2]*units('hPa') Pressure = df.iloc[:,2]*units.hectopascal TempC = df.iloc[:,3]*units.celsius RH = df.iloc[:,4]*units.percent Lat = df.iloc[:,7]*units.degree_north Lon = df.iloc[:,8]*units.degrees_E # Calculate dew pt temp Td = TempC - ((100-RH/5.)) wnd_dir = df.iloc[:,5]*units('degree') wnd_spd = df.iloc[:,6]*units('knots') rad = 4.0*np.arctan(1)/180. u_new = -wnd_spd*np.sin(rad*wnd_dir) v_new = -wnd_spd*np.cos(rad*wnd_dir) # Convert parameters into units Pressure = units.Quantity(Pressure.to_numpy(),"hectopascals") TempC = units.Quantity(TempC.to_numpy(),"degC") Td = units.Quantity(Td.to_numpy(),"degC") print('Press, TempC, Td:', Pressure.dtype, TempC.dtype, Td.dtype) Press, TempC, Td: float64 float64 float64 # Change default to be better for skew-T fig = plt.figure(figsize=(9, 11)) # Initiate the skew-T plot type from MetPy class loaded earlier skew = SkewT(fig, rotation=45) # Plot the data using normal plotting functions, in this case using # log scaling in Y, as dictated by the typical meteorological plot skew.plot(Pressure, TempC, 'r') skew.plot(Pressure, Td, 'g') # Set some appropriate axes limits for x and y skew.ax.set_xlim(-30, 40) skew.ax.set_ylim(1020, 100) print(Pressure) # mask = Pressure * hectopascal >= 100 mask = Presssure * hectopascal # Add the relevant special lines to plot throughout the figure skew.plot_dry_adiabats(t0=np.arange(233, 533, 10) * units.K, alpha=0.25, color='orangered') skew.plot_moist_adiabats(t0=np.arange(233, 400, 5) * units.K, alpha=0.25, color='tab:green') skew.plot_barbs(Pressure[mask][::75], u_new[mask][::75], v_new[mask][::75]) # skew.plot_mixing_lines(Pressure=np.arange(1000, 99, -20) * units.hPa, # linestyle='dotted', color='tab:blue') # Add some descriptive titles SHIP = 'Ron Brown: ', Lat[1], Lon[1] month=int(month) day = int(day) year = int(year) dtg = month,'/',day,'/',year table = str.maketrans(dict.fromkeys("()")) # dtg.translate(table) plt.title(format(SHIP), loc='left') plt.title(format(dtg), loc='right') plt.xlabel('Temperature (C)') plt.ylabel('Pressure hPa') plt.show() ``` `[1024.3 1022.4 1022.2 ... 98.2 98.1 98.1] hectopascal` ```pytb NameError Traceback (most recent call last)~\AppData\Local\Temp\12/ipykernel_13020/989837325.py in <module> 16 print(Pressure) 17 # mask = Pressure * hectopascal >= 100---> 18 mask = Presssure * hectopascal 19 20 # Add the relevant special lines to plot throughout the figure NameError: name 'Presssure' is not defined ``` [image: image.png]

[DanielAdriaansen]

You have a typo. Fix the spelling of "pressure" and try again and see if it works.

[arunaskuciauskas]

Dan: I fixed the spelling after sending this last code to you. However, I still get an error message after trying to calculate for 'mask' Arunas Code: print(Pressure) mask = Pressure * hectopascal >= 100 Error: [1024.3 1022.4 1022.2 ... 98.2 98.1 98.1] hectopascal

…

---------------------------------------------------------------------------NameError Traceback (most recent call last)~\AppData\Local\Temp\12/ipykernel_13020/737877772.py in <module> 16 print(Pressure)---> 17 mask = Pressure * hectopascal >= 100 18 NameError: name 'hectopascal' is not defined Arunas Kuciauskas ***@***.*** 831-261-4342

On Mon, Jan 3, 2022 at 4:52 PM Dan Adriaansen ***@***.***> wrote: You have a typo. Fix the spelling of "pressure" and try again and see if it works. — Reply to this email directly, view it on GitHub <#2279 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AQL3QFMUPEZ7NGSBDO7G473UUJAD5ANCNFSM5K7BY3JQ> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you authored the thread.Message ID: ***@***.***>

[DanielAdriaansen]

Since you have already converted Pressure to have the units "hectopascal", you don't need to specify the units again. As you see from printing Pressure:

print(Pressure)
[1024.3 1022.4 1022.2 ... 98.2 98.1 98.1] hectopascal

It already has units of hectopascals. So I think you can just use your original code of:

mask = Pressure >= 100

Try that and see if it works.

[arunaskuciauskas]

Dan: Yes, I tried it without the unit, but I still received an error. I’ll start up again tomorrow morning, thanks

…

Sent from my iPhone

On Jan 3, 2022, at 6:03 PM, Dan Adriaansen ***@***.***> wrote:  Since you have already converted Pressure to have the units "hectopascal", you don't need to specify the units again. As you see from printing Pressure: print(Pressure) [1024.3 1022.4 1022.2 ... 98.2 98.1 98.1] hectopascal It already has units of hectopascals. So I think you can just use your original code of: mask = Pressure >= 100 Try that and see if it works. — Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android. You are receiving this because you authored the thread.

[DanielAdriaansen]

Sounds good. I will review your code sample closer also and see if I notice anything else.

[arunaskuciauskas]

Dan: Here's the latest that I tried... I am able to properly plot a Skew-T profile, BUT, when I try to insert the Pressure, TempC[0] and Td[0] as in the following lne of code: parcel_path = mpcalc.parcel_profile(Pressure, TempC[0], Td[0]) it comes back with the below. I need help in trying to provide the Pressure (with units of hectopascal) in reverse order. Any help? Thanks, Arunas

…

---------------------------------------------------------------------------InvalidSoundingError Traceback (most recent call last)~\AppData\Local\Temp\13/ipykernel_11964/2429310348.py in <module> 5 #parcel_path = mpcalc.parcel_profile(Pressure, TempC[0], Td[0]) 6 ----> 7 parcel_path = mpcalc.parcel_profile(Pressure, TempC[0], Td[0]) ~\Anaconda3\lib\site-packages\metpy\xarray.py in wrapper(*args, **kwargs) 1214 1215 # Evaluate inner calculation-> 1216 result = func(*bound_args.args, **bound_args.kwargs) 1217 1218 # Wrap output based on match and match_unit ~\Anaconda3\lib\site-packages\metpy\units.py in wrapper(*args, **kwargs) 244 'that the function is being called properly.\n') + msg 245 raise ValueError(msg)--> 246 return func(*args, **kwargs) 247 248 return wrapper ~\Anaconda3\lib\site-packages\metpy\calc\thermo.py in parcel_profile(pressure, temperature, dewpoint) 742 743 """--> 744 _, _, _, t_l, _, t_u = _parcel_profile_helper(pressure, temperature, dewpoint) 745 return concatenate((t_l, t_u)) 746 ~\Anaconda3\lib\site-packages\metpy\calc\thermo.py in _parcel_profile_helper(pressure, temperature, dewpoint) 899 Pressure does not decrease monotonically in your sounding. 900 Using scipy.signal.medfilt may fix this."""--> 901 raise InvalidSoundingError(msg) 902 903 # Find the LCL InvalidSoundingError: Pressure does not decrease monotonically in your sounding. Using scipy.signal.medfilt may fix this. Arunas Kuciauskas ***@***.*** 831-261-4342

On Mon, Jan 3, 2022 at 6:14 PM Dan Adriaansen ***@***.***> wrote: Sounds good. I will review your code sample closer also and see if I notice anything else. — Reply to this email directly, view it on GitHub <#2279 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AQL3QFKOMZ7NBYNTBXY36A3UUJJYTANCNFSM5K7BY3JQ> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you authored the thread.Message ID: ***@***.***>

[DanielAdriaansen]

Hi Arunas,

I am able to use your code to compute parcel_path without getting that error. Have you made any other modifications to your code since this reply:
#2279 (comment)

?

[arunaskuciauskas]

Dan: can you repeat your message? My Gmail server could not provide me with your content. It might be best to respond to my work email; ***@***.*** Thanks

…

Sent from my iPhone

On Jan 3, 2022, at 6:14 PM, Dan Adriaansen ***@***.***> wrote:  Sounds good. I will review your code sample closer also and see if I notice anything else. — Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android. You are receiving this because you authored the thread.

[arunaskuciauskas]

Dan:
Hope I'm doing this properly. Just found out that I can't attach ipynb (code) files.
Attached is the WORD version, along with the output and error message...
Arunas

Code.docx

[DanielAdriaansen]

I am able to use your code to compute parcel_path without getting that error. Have you made any other modifications to your code since this reply:
#2279 (comment)

?

[arunaskuciauskas]

Dan: I’m coming back from a doc visit. Yes, I made a few of what I thought were insignificant changes. I can provide you with my updated code shortly. Thanks Get Outlook for iOS<https://aka.ms/o0ukef>

…

________________________________ From: Dan Adriaansen ***@***.***> Sent: Tuesday, January 4, 2022 2:25:43 PM To: Unidata/MetPy ***@***.***> Cc: arunaskuciauskas ***@***.***>; Author ***@***.***> Subject: Re: [Unidata/MetPy] Help with Skew-t plots that require units in variables (Discussion #2279) I am able to use your code to compute parcel_path without getting that error. Have you made any other modifications to your code since this reply: #2279<#2279> (comment) ? — Reply to this email directly, view it on GitHub<#2279 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AQL3QFJXVKH24UAKUBLLRETUUNXWPANCNFSM5K7BY3JQ>. Triage notifications on the go with GitHub Mobile for iOS<https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android<https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you authored the thread.Message ID: ***@***.***>

[arunaskuciauskas]

Dan: I attached the code. But just to convince you my problem area: parcel_path = mpcalc.parcel_profile(Pressure, TempC[0], Td[0]) error:

…

---------------------------------------------------------------------------InvalidSoundingError Traceback (most recent call last)~\AppData\Local\Temp\13/ipykernel_11964/3710001079.py in <module> 1 print(P, T1, dewpt)----> 2 parcel_path = mpcalc.parcel_profile(Pressure, TempC[0], Td[0]) ~\Anaconda3\lib\site-packages\metpy\xarray.py in wrapper(*args, **kwargs) 1214 1215 # Evaluate inner calculation-> 1216 result = func(*bound_args.args, **bound_args.kwargs) 1217 1218 # Wrap output based on match and match_unit ~\Anaconda3\lib\site-packages\metpy\units.py in wrapper(*args, **kwargs) 244 'that the function is being called properly.\n') + msg 245 raise ValueError(msg)--> 246 return func(*args, **kwargs) 247 248 return wrapper ~\Anaconda3\lib\site-packages\metpy\calc\thermo.py in parcel_profile(pressure, temperature, dewpoint) 742 743 """--> 744 _, _, _, t_l, _, t_u = _parcel_profile_helper(pressure, temperature, dewpoint) 745 return concatenate((t_l, t_u)) 746 ~\Anaconda3\lib\site-packages\metpy\calc\thermo.py in _parcel_profile_helper(pressure, temperature, dewpoint) 899 Pressure does not decrease monotonically in your sounding. 900 Using scipy.signal.medfilt may fix this."""--> 901 raise InvalidSoundingError(msg) 902 903 # Find the LCL InvalidSoundingError: Pressure does not decrease monotonically in your sounding. Using scipy.signal.medfilt may fix this. Arunas Kuciauskas ***@***.*** 831-261-4342

On Tue, Jan 4, 2022 at 2:28 PM Arunas Kuciauskas ***@***.***> wrote: Dan: I’m coming back from a doc visit. Yes, I made a few of what I thought were insignificant changes. I can provide you with my updated code shortly. Thanks Get Outlook for iOS <https://aka.ms/o0ukef> ------------------------------ *From:* Dan Adriaansen ***@***.***> *Sent:* Tuesday, January 4, 2022 2:25:43 PM *To:* Unidata/MetPy ***@***.***> *Cc:* arunaskuciauskas ***@***.***>; Author < ***@***.***> *Subject:* Re: [Unidata/MetPy] Help with Skew-t plots that require units in variables (Discussion #2279) I am able to use your code to compute parcel_path without getting that error. Have you made any other modifications to your code since this reply: #2279 <#2279> (comment) ? — Reply to this email directly, view it on GitHub <#2279 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AQL3QFJXVKH24UAKUBLLRETUUNXWPANCNFSM5K7BY3JQ> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you authored the thread.Message ID: ***@***.***>

[dopplershift]

@arunaskuciauskas Just FYI this is still getting posted to the GitHub discussion question, but all attachements are stripped. I'd suggest either taking it fully off GitHub (you need to use proper email addresses), or make sure things are attached to the discussion. Right now everyone following this is getting spammed with the messages but is unable to help since we can't see the code.

[arunaskuciauskas]

ok

[arunaskuciauskas]

Ryan: Do you suggest that I provide the entire code in my email back to Dan. I’m still learning the ropes…sorry

[dopplershift]

My suggestion would be if you're going to use email, don't use the github emails that are being used right now when you reply in your email client--you'll need to exchange your proper email addresses.

If you're going to use GitHub, don't reply to emails, but use the GitHub web interface so you can properly format your code to make it much easier to read. GitHub has the bare minimum support that's good for quick replies, but isn't good for sharing things.

[arunaskuciauskas]

Ryan:
Never mind. I should just look at the github portal link after the email message.
Arunas

[arunaskuciauskas]

Ryan: Thanks, I'm still not sure where (link) in github (MetPy) that I should go to, i.e., Is there a particular protocol in doing back and forth messaging? Thanks, Arunas Arunas Kuciauskas ***@***.*** 831-261-4342

…

On Tue, Jan 4, 2022 at 4:11 PM Ryan May ***@***.***> wrote: My suggestion would be if you're going to use email, don't use the github emails that are being used right now when you reply in your email client--you'll need to exchange your proper email addresses. If you're going to use GitHub, don't reply to emails, but use the GitHub web interface so you can properly format your code <https://docs.github.com/en/github/writing-on-github/working-with-advanced-formatting/creating-and-highlighting-code-blocks> to make it much easier to read. GitHub has the bare minimum support that's good for quick replies, but isn't good for sharing things. — Reply to this email directly, view it on GitHub <#2279 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AQL3QFJJ7A2LX37F2FK5XGDUUOEEBANCNFSM5K7BY3JQ> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you were mentioned.Message ID: ***@***.***>

[DanielAdriaansen]

Hi Arunas,

Perhaps you are more familiar with e-mail list servs, where one typically corresponds via e-mail. GitHub does support replying via e-mail, however it does not allow for all the nice formatting and attachment options that exist via https://www.github.com. Like @dopplershift, I would recommend using a web browser and logging in using your GitHub credentials when using GitHub Discussions for seeking help. When you receive an e-mail that someone has replied, rather than replying in e-mail, click the link in the e-mail at the bottom that says "view it on GitHub" which should then take you directly to GitHub in your default web browser, or go to your web browser and navigate to this page and reply.

I would prefer to correspond via Discussions on GitHub, rather than e-mail directly since this is public and will benefit others if they come across it. Hopefully this will work for you.

I was able to view your code in the Word Document you attached. It is very difficult to assist, because the code you copied is from a Jupyter notebook which has "cells" of code that can be executed separately, or all together in sequence, so viewing it as plain text copied directly from Jupyter is difficult.

That being said, the array of Pressure values you pass to parcel_path() MUST be monotonically decreasing, as the error message says. This means that there can't even be two values identical next to each other, because technically that's not decreasing, it's constant. For example, looking at your Pressure array I see a difference value of 0.0 hPa around a pressure level of 136.7 hPa (and many more above this pressure level). This means two adjacent levels in your sounding have the same exact pressure, which parcel_path() will not accept.

The quickest and easiest way I would solve this is by simply just restricting the Pressure data to be > 150 hPa passed to parcel_path(). All of the differences at pressures > 150 hPa appear to be negative (upon a quick inspection), indicating monotonically decreasing values. Of course if you add the parcel path to your Skew-T plot, then pressures between 150 and 100 hPa won't have the parcel path since I see you are using 100 hPa as your upper bound, so you can adjust your Skew-T upper limit pressure to match your parcel path pressures.

You can try this code:

mask2 = P >= 150.0
parcel_path = mpcalc.parcel_profile(Pressure[mask2], TempC[0], Td[0])

There are other ways to approach this. For example, you may consider performing some data pre-processing (smoothing, filtering, etc...) on your data if you don't need the exceptionally high resolution of the data. I notice there are nearly 4000 points in this profile, which is extraordinarily high resolution.

[dopplershift]

Thanks for digging in @DanielAdriaansen. I'll note that the next release of MetPy will relax the monotonic requirement to also allow adjacent points to be equal. I'll note that for most data, points don't actually have equal points, but this is merely an artifact of rounding to 1 decimal place.

[arunaskuciauskas]

Thanks for your responses. I’ll work on it later today. Currently, I’m moving out of my old place. Arunas

…

Sent from my iPhone

On Jan 5, 2022, at 9:47 AM, Ryan May ***@***.***> wrote:  Thanks for digging in @DanielAdriaansen. I'll note that the next release of MetPy will relax the monotonic requirement to also allow adjacent points to be equal. I'll note that for most data, points don't actually have equal points, but this is merely an artifact of rounding to 1 decimal place. — Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android. You are receiving this because you were mentioned.

[arunaskuciauskas]

Thank you very much!!!
OK, so I took your suggestions and I successfully plotted a Skew-T with shading of CIN (in blue) and CAPE (not shown because it does not exist), see attachment. I also added the LCL dot (can it be larger?). One item that I would like to add is the height scale along the right y axis. I've looked at MetPy Mondays #153. Code:
d = pandas_dataframe_to_unit_arrays(df)
skew.ax.text(1.08, Pressure, round(Hgt, 0), transform=skew.ax.get_yaxis_transform(which='tick2'))

However, here's the error message:

NameError Traceback (most recent call last)
~\AppData\Local\Temp\14/ipykernel_17364/3598646821.py in
1 # df.head()
----> 2 d = pandas_dataframe_to_unit_arrays(df)
3 # skew.ax.text(1.08, Pressure, round(Hgt, 0), transform=skew.ax.get_yaxis_transform(which='tick2'))

NameError: name 'pandas_dataframe_to_unit_arrays' is not defined

At the top of my code, I do have:
import pandas as pd
import matplotlib.pyplot as plt
import metpy.calc as mpcalc
from metpy.plots import SkewT
from metpy.units import units
import numpy as np
import re

Any help with this?
Thanks again,
Arunas

[arunaskuciauskas]

One other item: How do I remove the junk within the titles "[(()',]" at the top left and right?

[DanielAdriaansen]

Hi Arunas,

Per the documentation here:
https://unidata.github.io/MetPy/latest/api/generated/metpy.units.pandas_dataframe_to_unit_arrays.html

The pandas_dataframe_to_unit_arrays() function is found within the units module of MetPy. So you have two options:

1. import pandas_dataframe_to_unit_arrays explicitly like: from metpy.units import pandas_dataframe_to_unit_arrays
2. Tell your Python code that the function is part of the units module when you call it like this: d = units.pandas_dataframe_to_unit_arrays(df)

This is more general Python programming knowledge, rather than MetPy specific so I'd encourage you to consider exploring more about the Python language at this point, and how Python packages and modules are typically constructed and how imports work more generally. Your second question is also more Python related than MetPy, but in the spirit of encouraging more Python, I will offer a suggestion to remove the "title junk".

It looks like the relevant code for those strings is all of this:

df = pd.read_csv(fn, sep ='\s+', skiprows=8, nrows = 3800)

df_title_date = df_title[49:60]
day = df_title_date[0:2]
month = df_title_date[3:5]
year = df_title_date[6:11]

# Add some descriptive titles
SHIP = 'Ron Brown: ', Lat[1], Lon[1]
month=int(month)
day = int(day)
year = int(year)

dtg = month,'/',day,'/',year
plt.title(format(SHIP), loc='left')
plt.title(format(dtg), loc='right')

Again, this is basic Python programming knowledge about strings so I encourage you to practice and research various string manipulation strategies within Python. I'm not entirely sure what the code you've written is doing, but you need to concatenate the string objects properly.

I would suggest some options like this:

SHIP = 'Ron Brown: '+'%2.2f'+' %3.2f' % (float(Lat[1]),float(Lon[1])
dtg = '%02d %02d %04d' % (int(month),int(day),int(year))

I recommend Googling "Python string formatting" for lots of resources and various ways to do this. If you're using Python 3.6+, "f-strings" is a better way than what I've shown here. There are lots of options.

Glad your Skew-T is working!